NEW YORK (GenomeWeb) – Researchers from Stanford University have demonstrated that a next-generation sequencing-based assay they developed to analyze circulating tumor DNA can identify subtypes of diffuse large B cell lymphoma and predict patient outcomes.
In a study published in Science Translational Medicine this week, the Stanford team showed that the assay, which is based on the Cancer Personalized Profiling by Deep Sequencing (CAPP-seq) method, could potentially be used to identify subtypes of diffuse B cell lymphoma and to detect relapse-causing residual disease earlier than traditional methods.
Roche, which gained the rights to commercialize the CAPP-seq technology last year when it acquired Stanford spinout CAPP Medical, did not respond to requests for comments on its commercialization plans by press time. At the Association for Molecular Pathology meeting this week, the company previewed two CAPP-Seq research-use-only assays, but did not say when they would be officially available to customers.
The Stanford group originally described its method in a study published in 2014, and an updated version in a paper published earlier this year. At a conference in March, it also discussed using the assay to analyze ctDNA in non-small cell lung cancer patients.
Max Diehn and Ash Alizadeh, the senior authors of this week's study, told GenomeWeb that the method is essentially the same as the one they described earlier this year but targeted a different set of genes that are more specific to lymphoma.
The results of this study have two major implications, they said. One major question the researchers wanted to address was whether their assay could identify when a patient's disease transformed from nonaggressive to aggressive. "Historically, the only way to identify this is by biopsying a patient and looking under the microscope to see if the cells have changed," Diehn said. But this is impractical to do at regular time points to monitor patients, and a noninvasive way of monitoring would be much preferable. "We found preliminary evidence that the mutational profile changes as the patient's disease undergoes transformation events," he said. The researchers are now testing this ability in a larger group of patients.
The second major finding, Alizadeh said, is that the assay could identify different molecular subtypes. One reason for focusing on diffuse large B cell lymphoma is that it is a clinically and biologically heterogeneous disease with biomarkers that can be used for risk stratification and drug selection. Current methods for identifying those markers, however, rely on the availability of fresh frozen tumor samples. In addition, there is a subset of patients who progress from low-grade follicular lymphoma to DLBCL and for whom better biomarkers are needed.
The DLBCL targeted assay included known lymphoma-related SNVs, indels, and breakpoints involved in gene fusions, as well as the immunoglobulin heavy-chain variable region and heavy-chain joining cluster.
The researchers analyzed samples from 92 patients, including 76 tumor biopsies from patients who had been diagnosed with DLBCL as well as 144 longitudinal plasma samples, 45 of which were collected prior to any treatment. They found mutations in all tumor biopsy samples and found ctDNA in all pre-treatment plasma samples. In 39 out of 45 pre-treatment plasma samples, the researchers were able to identify at least one mutation known to be present in the tumor. And when ctDNA concentrations were above five haploid genome equivalents, they identified a known mutation in all samples. They also identified 95 percent of translocations involving BCL2, BCL6, and MYC.
In addition, the researchers found high concordance in allele frequencies between the plasma and biopsy samples.
To test whether the CAPP-seq assay could monitor tumor mutations over time and in response to therapy, the researchers used it on three patients with progressive disease who had received a B cell receptor inhibitor, ibrutinib. In two of the three patients, the assay found emerging resistance mutations developing at between 60 and 120 days after starting therapy.
The researchers also further analyzed the 45 pre-treatment samples, comparing the CAPP-seq assay with more traditional tests used to determine whether a patient is at high risk for disease progression. They found that the levels of ctDNA in the pre-treatment samples strongly associated with tumor volume. In addition, higher ctDNA levels were associated with worse outcomes, suggesting that the assay could "complement traditional clinical indices and serve as an independent prognostic biomarker," the authors wrote.
One major potential application of the assay, which the researchers are pursuing further, is to identify early signs of relapse. In 11 patients who eventually relapsed, the assay detected ctDNA at the time of relapse, and importantly, it also detected ctDNA prior to relapse in eight of them. When compared to a standard test for identifying disease progression, the ctDNA assay detected residual disease in twice as many patients, and on average, two months earlier. In addition, the CAPP-seq assay did not detect ctDNA in 10 patients who had been disease free for at least 24 months after therapy, nor did it detect ctDNA in 24 healthy adult samples.
A second important application of the assay is for subtyping, which could replace current methods that rely on fresh frozen tumor biopsies. Alizadeh, who is also a lymphoma oncologist, said it can be difficult to confirm a patient's subtype, often because biopsies cannot be safely taken. A disease subtype "can have life changing ramifications for what therapy you might pick," he added.
The researchers first used the standard tool, a microarray-based gene expression test, to molecularly subtype 76 patients with tumor biopsies as either germinal B-cell like or non-germinal B-cell like. They then used their ctDNA-based test to make the same classifications and demonstrated 80 percent concordance.
Diehn said the researchers are now expanding their study to larger cohorts of lymphoma patients. "Ultimately, we'd like to get to the point where we can make treatment decisions in the context of clinical trials," he said.
Diehn added that the lymphoma-specific CAPP-seq assay would be different from other NGS-based ctDNA assays currently on the market, which are primarily pan-cancer tests used to molecularly profile the tumor. Those tests don't typically cover mutations that are the most informative for lymphoma, he said.